Giving feedback to households about their energy consumption has been seen by many as a necessity in order for households to reduce their energy consumption and lower their carbon footprint. Many studies have been made on how smart meters, that give feedback on the total consumption, are used and their effect on the consumption. However, fewer studies have been done on how disaggregated feedback is understood and acted upon. Smappee is an energy feedback system that monitors the household’s consumption and is able to detect appliances’ consumption with only 3 physical clamps at the incoming current, using non-intrusive load monitoring (NILM). NILM differentiate appliances based on their electrical signature which is generated by turning an appliance on and off. The users get the feedback through a smartphone application. This study attempts to answer the question “How do users understand and act upon disaggregated feedback in Smappee?” by doing a qualitative study based on contextual interviews made on 15 users of Smappee. The results of the interviews are then compared with previous studies within the field of smart meters and a conceptual model is being described of how users understand Smappee. The results show that disaggregated feedback in Smappee is perceived as too difficult to acquire. Users thought it was too hard to find which appliances represented in Smappee correspond to which physical appliances in the household. However, the users used the real-time total consumption in order to make an estimation of how much certain appliances consumed. This indicates that users are interested in disaggregated feedback if it is easily accessible. Users have, in general, a good understanding of how Smappee detects appliances. They understand that Smappee is measuring the incoming current and makes assumptions based on the increases and decreases. They did not use the disaggregated feedback, even though they understood how it worked. / Att ge feedback till hushåll om deras energikonsumtion har setts, av många, som en nödvändighet för att hushåll ska kunna reducera sin energikonsumtion och minska sina koldioxidutsläpp. Flera studier har gjorts om hur smarta mätare, som ger feedback om totalkonsumtionen, används och deras effekt på förbrukningen. Men färre studier har gjorts om hur uppdelad feedback förstås och hur användare agerar på detta. Smappee är ett energifeedbacksystem som mäter hushålls konsumtion och är kapabel att detektera apparaters konsumtion med endast 3 klamrar på den inkommande strömmen, genom att använda “non-intrusive load monitoring (NILM). NILM skiljer på apparater baserat på deras elektriska signatur som genereras av att sätta på och slå av apparater. Användarna får feedbacken i en Smartphone-applikation. Denna studie försöker besvara frågan: “Hur förstår och agerar användare på uppdelad feedback i Smappee?” genom att genomföra en kvalitativ studie baserad på kontextuella intervjuer genomförda med 15 användare av Smappee. Resultatet från intervjuerna är sedan jämförda med tidigare studier inom området smarta mätare och en konceptuell model beskrivs om hur användare förstår sig på Smappee. Resultatet visar att uppdelad feedback i Smappee uppfattas för svår att använda. Användarna upplevde att det var för svårt att finna vilken apparat i Smappee som korresponderar till en viss fysisk apparat i hushållet. Dock använde användarna realtidskonsumtionen för att skapa en uppskattning om hur mycket vissa apparater konsumerade. Detta indikerar att användarna hade ett intresse i uppdelad feedback om det var tillgängligt på ett lättare sätt. Användare har, i allmänhet, en god förståelse om hur Smappee detekterar apparater. De förstår att Smappee mäter inkommande ström och gör antaganden baserat på ökningar och minskningar. De använde inte uppdelad feedback även om de förstod hur det fungerade.

This thesis presents the development of techniques which enable appliance recognition in an Advanced Electricity Meter (AEM) to aid individuals reduce their domestic electricity consumption. The key aspect is to provide immediate and disaggregated information, down to appliance level, from a single point of measurement. Three sets of features including the short term time domain, time dependent finite state machine behaviour and time of day are identified by monitoring step changes in the power consumption of the home. Associated with each feature set is a membership which depicts the amount to which that feature set is representative of a particular appliance. These memberships are combined in a novel framework to effectively identify individual appliance state changes and hence appliance energy consumption. An innovative mechanism is developed for generating short term time domain memberships. Hierarchical and nearest neighbour clustering is used to train the AEM by generating appliance prototypes which contain an indication of typical parameters. From these prototypes probabilistic fuzzy memberships and possibilistic fuzzy typicalities are calculated for new data points which correspond to appliance state changes. These values are combined in a weighted geometric mean to produce novel memberships which are determined to be appropriate for the domestic model. A voltage independent feature space in the short term time domain is developed based on a model of the appliance’s electrical interface. The components within that interface are calculated and these, along with an indication of the appropriate model, form a novel feature set which is used to represent appliances. The techniques developed are verified with real data and are 99.8% accurate in a laboratory based classification in the short term time domain. The work presented in this thesis demonstrates the ability of the AEM to accurately track the energy consumption of individual appliances.